Sensor Data Map Verification With Novel Region Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital maps face challenges in maintaining accuracy due to missing or erroneous data, which can affect route guidance and vehicle autonomy, especially in highly automated vehicles, necessitating efficient processing of vast sensor data to identify regions requiring updates.
Innovation Solution
A method and apparatus that analyze sensor data to determine consistency with a model, using consistent data to verify the model and identify regions needing updates, while filtering out redundant data to improve processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If all sensor data is processed to update digital maps, then map accuracy is improved, but processing time and computational resources increase significantly
Solution Approach 1:
The patent extracts only the novel portions of sensor data that differ from the existing digital map model, separating this novel data from redundant same data. This extraction approach allows the system to process only the necessary subset of data for map updates, significantly reducing processing time while maintaining map accuracy improvements.
Solution Approach 2:
The patent segments sensor data into distinct categories: novel data that requires processing for map updates, and same data that can be filtered out. This segmentation enables selective processing of only the relevant data portions, resolving the contradiction between comprehensive map accuracy improvement and excessive processing time.
2Reliability
If redundant sensor data is processed, then map verification is thorough, but processing efficiency decreases
Solution Approach 1:
The patent extracts and identifies same data that matches the existing digital map model, separating it from novel data. By extracting only the novel portions requiring processing, the system maintains thorough verification of map accuracy while dramatically improving processing efficiency by excluding redundant data from the processing pipeline.
Solution Approach 2:
The patent applies partial action by processing only the necessary novel data portions rather than all sensor data. This partial processing approach maintains sufficient map verification thoroughness while optimizing processing efficiency by avoiding excessive processing of redundant same data.
3Productivity
If sensor data is filtered to identify novel regions, then processing costs are reduced, but risk of missing important map updates increases
Solution Approach 1:
The patent employs feedback mechanisms where sensor data is systematically compared against the existing digital map model to identify discrepancies. This feedback loop ensures that novel data representing actual map changes is reliably detected and processed, while same data is efficiently filtered out, thus reducing processing costs without compromising map update completeness.
Solution Approach 2:
The patent replaces exhaustive mechanical processing of all sensor data with a smarter comparison-based system that substitutes detailed analysis with efficient model matching. This substitution identifies novel regions requiring updates while filtering redundant data, achieving cost efficiency without missing important map changes.
Data Source
AI summary
A method is provided for using sensor data that is consistent with the model to verify the model, and using sensor data that is different from the model to identify regions of the model requiring updates. Methods include: receiving observation data associated with a geographic area; comparing the observation data with a model of the geographic area; in response to the observation data failing to correspond to the model of the geographic area, identifying the observation data as novel data; in response to the observation data corresponding to the model of the geographic area, identifying the observation data as same data; using the novel data to update a map of a map database to form an updated map; and for observation data identified as same data, maintaining the map of the map database without using the same data to update the map.